129 



LYMANTRIIDAE 



Olene vagans B. & McD. (PI. XX, Figs. 8, 9). 



In the summer of 1917 whilst on a collecting trip in Maine, Dr. 

 McDunnough took two Olenc larvae from an apple tree at Packard's 

 Camp, Sebec Lake; at the same place on a hazel bush he also took a 

 single larva which was distinct from the preceding although closely 

 related. In the course of time the first two hatched into a S and 9 

 of the above species; the other larva produced a S (PI. XX, Fig. 7) 

 which on returning home and comparing with the material in the 

 collection we decided must be willingi which we had treated in our 

 revision as a race of vagans but which in the light of the larval dis- 

 tinctions must be regarded as a distinct species. 



Numerous S specimens of Olene taken at light at Sebec Lake 

 were carefully worked over and separated, using the bred material 

 as a basis; we found that with the exception of rubbed specimens 

 (which are very difficult to place correctly) the majority of 

 the S specimens could be fairly readily referred to either the 

 one form or the other by taking into account a combination of 

 the following features: (1) vagans has a rather darker color with 

 the cross lines more evident and with a tendency for the t. a. and t. p. 

 lines to approach each other quite closely at the inner margin; (2) 

 the dark band following the t. p. line in vagans is quite well defined 

 outwardly by an irregular white s. t. line terminating in a distinct white 

 spot above inner margin and showing slight dark transverse dashes 

 below costa; in willingi the s. t. area is very poorly defined and the 

 white spot is far less noticeable; (3) the reniform is more distinctly 

 white shaded in vagans than in willingi. 



With regard to the 9 's we are unable to give any comparisons, 

 as the only 9 secured was the one bred from apple; we beheve 

 however on the strength of this specimen that our 9 types of vagans 

 belong to another species and that the name must be restricted to 

 the $ type figured in our Contributions, Vol. II, (2) PI. Ill, Fig. 1 ; 

 we figure the correct 9 (PI. XX, Fig. 9), and a comparison of 

 this figure with those on PI. Ill, Figs. 2 and 4 of the revision will 

 show the marked diflferences; it may be that our 9 Co-type should 



